Improving automated test case generation with PSO and MCMC: A pattern recognition approach in software testing

High-performance contexts require sophisticated testing techniques to ensure reliable performance in increasingly complex software systems. In most cases, traditional defect detection techniques are unable to stay effective. This paper proposes a hybrid approach to overcome these limitations by combining Markov Chain Monte Carlo (MCMC) and Particle Swarm Optimisation (PSO) methods. The proposed method merges the exploration powers of PSO with the effective probabilistic sampling of MCMC in the paradigm of hybrid optimisation. This reduces computational redundancy and cost, while improving the process of fault identification and test coverage in high-dimensional testing settings. The main objectives, especially for complex, high-performance software systems, are to decrease computational redundancy, increase test coverage, improve error identification, and increase fault detection rates. The hybrid PSO-MCMC model achieved 85% test coverage, 15% redundancy reduction, and 93% fault detection accuracy. Earlier approaches such as NLP-based strategies and Deep Variational Autoencoding Models are examples of earlier approaches that do not meet these results. The results indicate that the PSO-MCMC hybrid model is a promising approach for complex systems since it enhances automated software testing by reducing redundancy and increasing fault identification and test coverage.